The Contractor intends to select randomly 100 out of about 1000 imprint specimens from different tumors. The specimens will be interactively analyzed in their cell anlysis sytem and the result from this processing will be saved and later subjected to careful statistical analysis on a large-scale computer, in order to define a measure characterizing nuclear grade for each one of the selected tumors. He also intends to see if there are any natural categorization of the tumors from the nuclear grade point of view. The results from these experiments will be compared with the conventional classification used by pathologists and cytologists. In order to evaluate the importance of the obtained measure of nuclear grading he intends to relate it to the three and five year survival of the patients in a follow-up study.